Cambridge study: 72-92% of submarine fiber optic cables would need to be cut simultaneously to significantly impact the Bitcoin network.

PANews reported on March 14th that, according to CoinDesk, a study released by the Cambridge Centre for Alternative Finance (CCAF) shows that the Bitcoin network is far more resilient to physical infrastructure disruptions than previously expected. The study, based on 11 years of network data and analysis of 68 verified submarine cable failures, found that in the event of a random failure, 72% to 92% of transnational submarine cables would need to be interrupted simultaneously to significantly impact Bitcoin node connectivity.

The study indicates that in over 87% of the 68 real-world submarine cable failures analyzed, the impact on nodes was less than 5%. For example, in March 2024, submarine disturbances off the coast of Côte d'Ivoire damaged 7 to 8 cables simultaneously, affecting approximately 43% of nodes in the region, but globally only about 5 to 7 Bitcoin nodes were affected, representing approximately 0.03% of the network. The Bitcoin network is significantly more vulnerable to targeted attacks than to random failures. If an attacker targets a critical communication hub's cable, impacting only about 20% of the critical lines could cause a similar disruption; a coordinated disruption targeting the hosting providers with the most nodes (such as Hetzner, OVH, Comcast, Amazon, and Google Cloud) could result in a significant impact with a reduction in routing capacity of about 5%.

Furthermore, the study found that Tor network nodes accounted for approximately 64% (in 2025), with their distribution mainly concentrated in highly connected countries such as Germany, France, and the Netherlands. This, in turn, improved the overall resilience of the network. The study suggests that the Bitcoin community is gradually shifting towards more censorship-resistant infrastructure in the face of internet censorship and geopolitical events, giving the network an "adaptive reinforcement" characteristic.